Summary

In this chapter, you used a variety of TensorFlow resources, including TensorBoard, TensorFlow Hub, and Google Colab. TensorBoard offers users a method to visualize computational model graphs, metrics, and any experimentation results. TensorFlow Hub allows users to accelerate their machine learning development using pre-trained models built by experts in the field. Google Colab provides a collaborative environment to develop machine learning models on Google servers. Developing performant machine learning models is an iterative process of trial and error, and the ability to visualize every step of the process can help practitioners debug and improve their models. Moreover, understanding how experts in the field have built their models and being able to utilize the pre-learned weights in the networks can drastically reduce training time. All of these resources are used to provide an environment to develop and debug machine learning algorithms in an efficient workflow.

...